Enzyme Replacement Therapy in CLN2-Associated Retinopathy

Keywords

genetics - pediatric ophthalmology - retina

Schlüsselwörter

Genetik - Kinderophthalmologie - Retina

Introduction

In the 1990 s, intravenous imiglucerase was launched as the first enzyme replacement therapy (ERT) for the treatment of the lysosomal storage disorder (LSD) Gaucher disease. Since then, intravenous ERT has become the gold standard for seven further LSDs: Pompe disease, Fabry disease, and mucopolysaccharidoses (MPSs) I, II, IV A, VI, and VII, and the number is still growing [1].

In these disorders, ERT is based on the intravenous administration of a recombinant human enzyme exploiting the mechanism of lysosomal enzyme recycling and biogenesis. For lysosomal targeting, the recombinant soluble hydrolases are tagged with a mannose-6-phosphate (M6P) group linked to N-linked oligosaccharides, as it occurs in vivo when the endogenous enzymes are synthesized in the endoplasmic reticulum and pass through the cis-Golgi network. Proteins fused to MP6 groups are recognized by MP6 receptors present in the trans-Golgi network and packed into clathrin-coated vesicles that ultimately become mature lysosomes. This natural feature of lysosomal biogenesis enables for a targeted delivery of recombinant enzymes to lysosomes and makes ERT an elegant tool to treat LSDs caused by enzyme deficiencies [2].

Intravenous ERT slows disease progression and improves visceral clinical symptoms. For example, ERT in MPSs I, II, VI, and IV effectively reduces liver and spleen volumes and the excretion of urinary glycosaminoglycans (for review see [1]). However, bradytrophic tissues such as the cornea, cartilage, or bone are poorly reached by ERT. In addition, the effect on the central nervous system and retinal structures is negligible due to the blood-brain and blood-retinal barriers [2], [3].

To overcome these limitations, intracerebroventricular routes of administration have successfully been established for two LSDs with a primary neurodegenerative course, namely, neuronal ceroid lipofuscinosis type 2 (CLN2) and Sanfilippo syndrome type B. For the ERT of these conditions, an Ommaya or Rickham ventricular reservoir is neurosurgically implanted under the scalp, with a catheter placed in the cerebral lateral ventricle. ERT administered via these access routes has been shown to slow neurodegeneration in both diseases. While the intracerebroventricular ERT of CLN2 disease with cerliponase alfa was already approved by the EMA and FDA in 2017, tralesinidase alfa for the treatment of Sanfilippo syndrome type B is still subject of clinical trials (NCT02754076) [4], [5].

Neuronal Ceroid Lipofuscinosis Type 2 (CLN2)

Neuronal ceroid lipofuscinosis type 2 (CLN2 disease; OMIM#204500) belongs to the spectrum of neuronal ceroid lipofuscinoses (NCLs). NCLs are inherited LSDs clinically characterized by neurodegeneration, resulting in cognitive and motor regression, epilepsy, myoclonus, ataxia, loss of vision, and shortened life expectancy. NCLs arise from pathogenic variants in 1 of 13 known genes that code mainly for soluble lysosomal hydrolases or transmembrane proteins. The age at onset of symptoms is typically during childhood but varies between NCL types. Almost all patients appear healthy at birth and show a period of normal development [6], [7]. The sequence of symptoms is variable and depends on the age of onset and CLN type [8]. As a group, NCLs represent the most common pediatric neurodegenerative disease.

CLN2 disease results from biallelic pathogenic variants in the TPP1 gene encoding the lysosomal enzyme tripeptidyl peptidase (TPP1). TPP1 is synthesized as an inactive 66 kDa precursor that is autocatalytically converted to an active 46 kDa serine protease at an acidic pH. Active TPP1 cleaves tripeptides from the amino terminus of small polypeptides marked for degradation. Its deficiency results in an accumulation of cytotoxic lipofuscein-containing storage material in lysosomes of neuronal tissue, including the retinal cells of the eye [9]. More than 50% of patients are homozygous or compound heterozygous for the TPP1 variants c.622C>T (p. Arg208*) or c.509 – 1G>C. The latter is associated with a particularly severe course of CLN2 retinopathy [10].

The classic late-infantile phenotype of CLN2 disease manifests between 2 and 4 years of age with an unprovoked seizure, ataxia, or myoclonus. Delayed speech development is characteristic for this CLN type and may even precede the occurrence of neurological symptoms [11]. Neurological decline in CLN2 disease follows a largely predictable course. The initial symptoms are soon followed by an accelerated loss in cognitive, language, and motor functions from the age of 36 to 60 months, with complete loss of motor, speech, and visual functions by 6 and 7 years of age. Later in the disease course, intractable epilepsy, progressive dementia, and the loss of motor functions, including the ability to swallow, lead to premature death, usually between 8 to 12 years. To monitor disease progression, main neurological symptoms are assessed by a two-domain motor and language CLN2 disease rating scale [10], [12].

CLN2 Retinopathy

In contrast to most other NCL types, loss of vision in the natural course of CLN2 occurs when progressive mental regression and gross motor disturbances have already become apparent. In untreated cohorts, the average onset of obvious visual symptoms was reported at an age of 48 months. Loss of visual functions then follows a tilted S-shaped decline, with rapid progression between 48 and 60 months leading to blindness within 3 years [13], [14].

Because of the already advanced psychomotor decline, previous ophthalmologic scoring systems had to rely on a rough visual behavioral scale without eye exams or ancillary testing when loss of visual functions become evident. In that scoring system, visual functions were given four different scores: recognizing a desired object and grabbing at it was rated as 3, grabbing for objects but uncoordinated as 2, reacting to light as 1, and no reaction to visual stimuli as 0 [12].

More recent studies demonstrate that CLN2 retinopathy functionally resembles a primary cone-rod dystrophy initially presenting as a maculopathy, with the first identifiable alterations perceived by optical coherence tomography (OCT) in the parafovea between 39 and 48 months ([Fig. 1 a]) [7], [13], [14]. The retinopathy shows a remarkable symmetry and is now classified in a scoring system based on fundus appearance and OCT findings (ophthalmic severity score; Weill Cornell LINCL Ophthalmic Severity Scale, Weill Cornel Batten Score; WCBS) [13], [14]. The score divides ocular findings into five severity categories: (1) score 1, representing a normal fundus, (2) score 2, pigmentary changes in the parafoveal region (average age: 58.6 months) with parafoveal disruption of the ellipsoid band but intact external limiting membrane (ELM) in OCT, (3) score 3, which denotes bullʼs eye maculopathy, with outer retinal and photoreceptor loss as well as the ELM confined to the central fovea (extending less than 1 disc diameter; average age 69 months), (4) score 4, a more extensive degeneration, with a bullʼs eye pattern extending more than one disc diameter from the fovea and outer retinal atrophy in OCT that extends less than 2 disc diameters with normal appearing retina beyond the central fovea (mean age 61 months), and (5) score 5, a bullʼs eye maculopathy greater than 2 disc diameters, which equals to diffuse outer retinal atrophy in OCT at a mean age of 81 months [13], [14]. To date, these anatomical data have not been correlated with visual acuity, with the exception of one case report [15]. This is largely attributable to the fact that in most studies, data were acquired in general anesthesia and the advanced neurological stage of the patients precluded reproducible best-corrected visual acuity (BCVA) testing.

Fig. 1 Spectral-domain optical coherence tomography (SD-OCT) and near-infrared (NIR) images showing the delayed progression of CLN2-associated cone-rod dystrophy in a child treated with intravitreal enzyme replacement with cerliponase alfa in the right eye (OD), while the left eye (OS) remained untreated. a At the age of 49 months, parafoveal disruption of the ellipsoid band is apparent (WCBS2). b At the beginning of treatment, at an age of 60 months, bullʼs eye maculopathy is evident, both in SD-OCT and NIR images, equaling to a WCBS3. BCVA was RE/LA 0.5. c OD: The arrows depict the transition between the preserved outer retinal layers and the disappearance of ellipsoid zone, external limiting membrane, and outer nuclear layer in the sub- and perifoveal area (WCBS4). OS: Diffuse and pronounced loss of the outer retinal layers in the entire scan area with concomitant choroidal hypertransmission exemplary for a concomitant rarefaction of the retinal pigment epithelium (WCBS5). The more extensive area of retinal atrophy in the untreated OD is also recognizable in the NIR images.

Intracerebroventricular Enzyme Replacement Therapy for CLN2 Disease

In 2017, cerliponase alfa (Brineura, BioMarin Pharmaceutical Inc., Novato CA, USA), a recombinant human TPP1 (rhTPP1) enzyme, was approved for intracerebroventricular ERT. Cerliponase alfa is administered as an intracerebroventricular infusion over several hours every 2 weeks using an Ommaya or Rickham device as described above [4]. Patients treated with cerliponase alfa have been found to show a markedly slower cognitive decline and preserved motor and language function compared to untreated historical controls [4], [16]. Up to now, CLN2 disease is the only neurodegenerative LSD treatable by an (intracerebroventricular) ERT. As shown in a recent case report, treatment started at a presymptomatic stage even has the potential to delay disease onset, which is currently being investigated in a phase 2 trial (NCT02678689) [9].

Progression of the CLN2 retinopathy, however, remains unaffected, most likely because the intracerebroventricular enzyme does not cross the blood-retinal-brain barrier [4], [17], [18], [19]. This contrasts post-retinal visual pathway neurodegeneration in CLN2 disease, which appears to be influenced by the treatment. For example, flash visual evoked potentials (VEPs), which reflect cortical visual function, were found to be absent or delayed in treatment-naive patients. In contrast, the majority of patients who received intracerebroventricular ERT had preserved VEPs, albeit with an abnormal early photopic paroxysmal response, suggesting some stage of neurodegeneration [7], [18], [20], [21]. This is reasonable, since the optic nerve is surrounded by cerebrospinal fluid (CSF) circulating with in the subarachnoid space. However, the lamina cribrosa, a structural element of the optic nerve, represents a barrier from the intraocular space [22]. In addition, intracerebroventricular administration does not achieve pharmacologically effective concentrations in the serum that could benefit the retina. Pharmacokinetic studies showed that rhTPP1 peaks in the CSF 4 hours after intracerebral application and after 8 hours in the plasma. The respective plasma levels are 300- to 1000-fold lower than in the CSF, suggesting blood-brain barrier leakage. Since the retina is shielded from blood circulation by the inner and outer blood-retinal barrier, represented by the non-fenestrated retinal capillaries and the Bruchʼs membrane/retinal pigment epithelium complex, it is unlikely that therapeutic levels are reached within the eye [23].

As cognitive, neurological, and language decline are markedly delayed by intracerebroventricular ERT, the isolated and rapid loss of visual functions becomes a huge burden for the patients and their families. The loss of vision is actively perceived. This is especially true for children who were able to start treatment early.

Intravitreal Enzyme Replacement Therapy to Delay the Clinical Course of CLN2 Retinopathy: Early Experience

An intravitreal (IVT) enzyme replacement strategy to directly reach the target tissue was conceivable considering the clinical routines of the various IVT anti-VEGF therapies and the fact that, with cerliponase alfa, an effective pharmaceutical preparation was already available.

This concept was first tested in a canine model of CLN2. In TPP1-null dogs, IVT administration of rhTPP1 was initiated at an age of 12 weeks before any signs of retinal degeneration or loss of function were detectable. Treatment was administered every 2 weeks with a dose up to 0.3 mg until the dogs had reached end-stage neurological decline, between 43 and 46 weeks of age. All dogs exhibited preservation of rod and cone b-wave amplitudes in the treated eye as assessed by electroretinography. Electron microscopic examination indicated that distal ends of the outer segments remained tightly packed and straight in the treated eye, while gaps between the discs were present in the vehicle-treated eyes. Nevertheless, intraocular inflammation was evident in all treated eyes, both clinically and histologically [24]. In the same animal model, an IVT injection of autologous mesenchymal stem cells (MSCs) programmed to overexpress and secrete human TPP1 was also effective in preserving retinal function and structure without eliciting an inflammatory response [25]. Whereas the MSC approach has not yet been translated to humans, the promising results from IVT administration of rhTPP1 in dogs paved the way for the first treatments in patients with CLN2 disease.

First results on feasibility, safety, and efficacy of IVT-ERT with cerliponase alfa for CLN2 retinopathy came from a recent prospective, interventional, controlled, single-center, compassionate-use study conducted at the Great Ormond Street Hospital in London. The study enrolled eight severely affected patients at a median age of 7.5 years. At baseline, the mean ophthalmic score was 4, suggestive of advanced retinopathy. Patients received 0.2 mg rhTPP1 in 0.05 mL BSS every 8 weeks into the right eye under general anesthesia for 12 to 18 months. The left eye served as an untreated control. IVT-ERT was prepared from the cerliponase alfa overage of the same patientʼs intracerebroventricular treatment. The dose was chosen from scaling up the 0.1 mg from the canine model assuming that the human vitreous volume is approximately twice that of a dog [24], [26], [27]. Since uveitis was known from the canine model, dexamethasone eye drops were prescribed for 4 weeks after each injection. The primary outcome was safety, and the secondary outcome was efficacy. There were no events of uveitis, raised intraocular pressure, or iatrogenic ocular pathology. Transient central artery occlusions in two patients were relieved by paracentesis and one patient experienced laryngospasm from general anesthesia. Since the study was focused on safety, five patients already had end-stage retinopathy at enrollment. In these subjects, serial OCT scans did not reveal differences between the treated and untreated eye. However, in patients with less advanced disease (n = 3) who still exhibited a bilateral decline in paramacular volume as measured by supine macular OCT, the mean rate of decline was slower in the treated eyes as compared to the untreated eyes, clearly demonstrating that IVT-ERT was effective in reducing the rate of macular volume loss, although all of them were in the actively degenerating phase of the disease [26]. Another ongoing clinical phase I/II study is investigating the safety and efficacy of IVT rhTPP1 at 4-week intervals over 24 months (NCT05152914). The results have not yet been published. Encouraged by the promising results from the Great Ormond Street Hospital, several European centers are currently starting individual treatment with IVT-ERT. Dosing and treatment frequencies of the respective departments are provided in [Table 1].

Table 1 Intravitreal enzyme replacement therapy in European and North American centers.

Country	# of participants	Dose	Cycle	Comments
IVT-ERT: intravitreal enzyme replacement therapy
Europe
Germany, Munich	1	0.4 mg (0.05 mL)	4-weekly	until IVT-ERT #7 0.2 mg (0.05 mL)
Netherlands, Rotterdam	2	0.4 mg (0.05 mL)	6-weekly
United Kingdom, London	8	0.2 mg (0.05 mL)	8-weekly	Wawrzynski et al., 2024
Spain, Córdoba	1	0.2 mg (0.05 mL)	8-weekly	both eyes
North America
Ohio, United States	5	0.2 mg (0.02 mL)	4-weekly	after #12 IVT-ERT both eyes; NCT05152914
Canada, Toronto	1	0.4 mg (0.05 mL)	4-weekly	planned dose increase to 0.6 mg

Our own experience from an individual treatment confirms both safety and efficacy of IVT-ERT with cerliponase alfa ([Fig. 1 c]). In this individual treatment, a patient homozygous for the known pathogenic variant c.509 – 1 G>C in the TPP1 gene, who was started on intracerebroventricular ERT as early as the age of 40 months, received IVT-ERT in the right eye (OD) every 4 weeks since the age 60 months, while the left eye (OS) served as a paired untreated control. After 7 IVT injections, the initial dosage of 0.2 mg rhTPP1 in 0.05 mL BSS was increased to 0.4 mg in 0.05 mL. Paracentesis for occlusion of the central retinal artery did not have to be performed. Ophthalmological exams 3 days after injection were unremarkable, with no signs of intraocular inflammation.

One year before the first IVT-ERT with cerliponase alfa, at the age of 49 months, our patient had a CLN2 ophthalmic severity score of 2 ([Fig. 1 a]). Within the 11 months until the start of IVT-ERT retinopathy, he progressed to a score of 3 in both eyes, suggesting that he was in the active degenerating phase of the disease ([Fig. 1 b]). The dynamics in retinopathy progression are consistent with results from natural history cohorts. The severity of the retinopathy in our patient, however, underscores the severe ocular phenotype of the underlying c.509 – 1G>C TPP1 variant. For comparison, in a cohort of patients averaging multiple genotypes, an ophthalmic severity score of 2 was only reached at an average age of 58.9 months, a score of 3 at 69.0 months, and a score of 5 at 81.0 months [13].

Initially, CLN2 retinopathy continued to progress rather symmetrically in both eyes for 6 – 7 months. Such a delay in the influence on structural neurodegeneration is also known from intracerebroventricular cerliponase alfa treatment [4] and was also observed in the British prospective, interventional, controlled, compassionate-use study in those four participants, who were assumed to still be in the actively degenerating phase (score 3 – 4) [26], [27]. Thereafter, a functional and structural difference between the treated and the paired control eye became apparent. In more detail, 5 months after the first IVT-ERT, BCVA had declined from OD/OS0.5 to OD 0.25, OS0.32 decimal. As shown in [Fig. 1 c], the untreated OS continued to deteriorate, as predicted from the natural course (BCVA after 15 months ERT: OD 0.2; OS0.1), while the treated RE began to stabilize. In line with this, after 24 months of IVT-ERT, at an age of 6.5 years, OD still had a WCBS of 4 and a BCVA of 0.2, whereas OS had progressed to stage 5 and a BCVA of counting fingers (clinical and ophthalmological details are reported in Priglinger CS et al. [15]).

An inter-eye difference was also conceivable from a behavioral point of view, owing to the early start of intracerebroventricular ERT, the remarkably compliant child became increasingly frustrated upon occlusion of the treated eye in visual acuity testing. This was also evident from the acquisition of OCT scans, which was already increasingly hampered from instable fixation 4 – 5 months after beginning treatment.

Conclusion

CLN2 disease is the first inherited retinal dystrophy that may benefit from IVT-ERT, most notably if begun before the first signs of photoreceptor involvement are evident. First case reports show that IVT cerliponase alfa not only slows down anatomical retinal degeneration but is also associated with a transient stabilization of visual acuity. This is highly relevant, because patients undergoing intracerebroventricular ERT with cerliponase alfa experience a markedly slower cognitive decline. The younger they are at the initiation of treatment, the more they will inevitably have to experience a presumably treatable loss of visual function that in the natural course occurred after severe cognitive decline.

The major drawback of ERT is the short half-life of the recombinant enzyme which obligates a repetitive, frequent and lifelong administration, with the requirement of repetitive general anesthesia for IVT-ERT in children. Studies with larger study populations, like the ongoing and planned clinical trials are needed to optimize dosing and timing of treatment. Alternative delivery strategies such as sustained delivery devices or dose escalation strategies are warranted to avoid or reduce the frequency of general anesthesia required in pediatric patients for this adult outpatient setting surgical procedure. Nevertheless, IVT-ERT is a valuable option for those patients who will not be eligible for the planned gene therapy trial (NCT05791864) or due to a rapid course of retinopathy that will have progressed too far when gene therapy may become available.

• References

• 1 Fernández-Pereira C, San Millán-Tejado B, Gallardo-Gómez M. et al. Therapeutic Approaches in Lysosomal Storage Diseases. Biomolecules 2021; 11: 1775
• 2 Coutinho MF, Prata MJ, Alves S. A shortcut to the lysosome: the mannose-6-phosphate-independent pathway. Mol Genet Metab 2012; 107: 257-266
• 3 Nagpal R, Goyal RB, Priyadarshini K. et al. Mucopolysaccharidosis: A broad review. Indian J Ophthalmol 2022; 70: 2249-2261
• 4 Schulz A, Ajayi T, Specchio N. et al. Study of Intraventricular Cerliponase Alfa for CLN2 Disease. N Engl J Med 2018; 378: 1898-1907
• 5 Muschol N, Koehn A, von Cossel K. et al. A phase I/II study on intracerebroventricular tralesinidase alfa in patients with Sanfilippo syndrome type B. J Clin Invest 2023; 133: e165076
• 6 Specchio N, Pietrafusa N, Trivisano M. Changing Times for CLN2 Disease: The Era of Enzyme Replacement Therapy. Ther Clin Risk Manag 2020; 16: 213-222
• 7 Thompson DA, Handley SE, Henderson RH. et al. An ERG and OCT study of neuronal ceroid lipofuscinosis CLN2 Battens retinopathy. Eye (Lond) 2021; 35: 2438-2448
• 8 Goebel HH, Schochet SS, Jaynes M. et al. Progress in neuropathology of the neuronal ceroid lipofuscinoses. Mol Genet Metab 1999; 66: 367-372
• 9 Schaefers J, van der Giessen LJ, Klees C. et al. Presymptomatic treatment of classic late-infantile neuronal ceroid lipofuscinosis with cerliponase alfa. Orphanet J Rare Dis 2021; 16: 221
• 10 Nickel M, Simonati A, Jacoby D. et al. Disease characteristics and progression in patients with late-infantile neuronal ceroid lipofuscinosis type 2 (CLN2) disease: an observational cohort study. Lancet Child Adolesc Health 2018; 2: 582-590
• 11 Simonati A, Williams RE. Neuronal Ceroid Lipofuscinosis: The Multifaceted Approach to the Clinical Issues, an Overview. Front Neurol 2022; 13: 811686
• 12 Steinfeld R, Heim P, von Gregory H. et al. Late infantile neuronal ceroid lipofuscinosis: quantitative description of the clinical course in patients with CLN2 mutations. Am J Med Genet 2002; 112: 347-354
• 13 Orlin A, Sondhi D, Witmer MT. et al. Spectrum of ocular manifestations in CLN2-associated batten (Jansky-Bielschowsky) disease correlate with advancing age and deteriorating neurological function. PLoS One 2013; 8: e73128
• 14 Kovacs KD, Patel S, Orlin A. et al. Symmetric Age Association of Retinal Degeneration in Patients with CLN2-Associated Batten Disease. Ophthalmol Retina 2020; 4: 728-736
• 15 Priglinger CS, Courage C, Lotz-Havla AS. et al. Intravitreal enzyme replacement therapy slows retinopathy in late infantile ceroid lipofuscinosis type 2. Neuropediatrics 2025; (epub ahead of print)
• 16 Schulz A, Specchio N, de Los Reyes E. et al. Safety and efficacy of cerliponase alfa in children with neuronal ceroid lipofuscinosis type 2 (CLN2 disease): an open-label extension study. Lancet Neurol 2024; 23: 60-70
• 17 Estublier B, Cano A, Hoebeke C. et al. Cerliponase alfa changes the natural history of children with neuronal ceroid lipofuscinosis type 2: The first French cohort. Eur J Paediatr Neurol 2021; 30: 17-21
• 18 Quagliato EMAB, Rocha DM, Sacai PY. et al. Retinal function in patients with the neuronal ceroid lipofuscinosis phenotype. Arq Bras Oftalmol 2017; 80: 215-219
• 19 Dulz S, Schwering C, Wildner J. et al. Ongoing retinal degeneration despite intraventricular enzyme replacement therapy with cerliponase alfa in late-infantile neuronal ceroid lipofuscinosis type 2 (CLN2 disease). Br J Ophthalmol 2023; 107: 1478-1483
• 20 Horiguchi M, Miyake Y. Batten disease–deteriorating course of ocular findings. Jpn J Ophthalmol 1992; 36: 91-96
• 21 Jadav RH, Sinha S, Yasha TC. et al. Clinical, electrophysiological, imaging, and ultrastructural description in 68 patients with neuronal ceroid lipofuscinoses and its subtypes. Pediatr Neurol 2014; 50: 85-95
• 22 Mirra S, Marfany G, Garcia-Fernandez J. Under pressure: Cerebrospinal fluid contribution to the physiological homeostasis of the eye. Semin Cell Dev Biol 2020; 102: 40-47
• 23 OʼLeary F, Campbell M. The blood-retina barrier in health and disease. FEBS J 2023; 290: 878-891
• 24 Whiting REH, Robinson Kick G, Ota-Kuroki J. et al. Intravitreal enzyme replacement inhibits progression of retinal degeneration in canine CLN2 neuronal ceroid lipofuscinosis. Exp Eye Res 2020; 198: 108135
• 25 Tracy CJ, Whiting RE, Pearce JW. et al. Intravitreal implantation of TPP1-transduced stem cells delays retinal degeneration in canine CLN2 neuronal ceroid lipofuscinosis. Exp Eye Res 2016; 152: 77-87
• 26 Wawrzynski J, Martinez AR, Thompson DA. et al. First in man study of intravitreal tripeptidyl peptidase 1 for CLN2 retinopathy. Eye (Lond) 2024; 38: 1176-1182
• 27 Rodriguez-Martinez AC, Wawrzynski J, Henderson RH. Intravitreal enzyme replacement for inherited retinal diseases. Curr Opin Ophthalmol 2024; 35: 232-237